The .web gTLD could go live in 2016

The new gTLD .web could be coming to the internet sooner than expected after two of the remaining barriers to delegation disappeared.
Following the withdrawal last week of an application for the plural .webs, an auction for .web could happen in the next couple of months, enabling a go-live date possibly in 2016.
.web, often considered the most desirable truly generic gTLD, has had a rough time of it in the 2012 ICANN new gTLD program.
There were seven applications for the string. Google, Web.com, Donuts, Radix, Afilias, Schlund Technologies, Nu Dot Co all applied.
The registrar Web.com (owner of Network Solutions, Register.com, et al) appears to be especially keen to get the domain, given that the string more or less matches its brand.
It perhaps should have been a straightforward auction shoot-out.
But, complicating matters, bespoke printing firm Vistaprint had filed two applications — one vanilla, one “community” based — for the plural version of the string, “.webs”.
Vistaprint runs a website development service called Webs.com. It’s the plural of the Web.com brand.
Web.com wasn’t happy about Vistaprint’s .webs applications, so it filed String Confusion Objections against both, arguing that .web and .webs were too confusingly similar to co-exist on the internet.
While there are now many examples of plurals and singulars living together (see .auto/s, .fan/s and .gift/s), the registrar won both of its SCO complaints, meaning Vistaprint’s two applications and the seven .web applicants were lumped together into the same contention set.
If two strings are in the same contention set, only one can survive to be ultimately delegated to the DNS.
Vistaprint appealed the SCO decisions, first with a Request for Reconsideration to the ICANN board (predictably unsuccessful) and then with an Independent Review Process complaint.
While the IRP was being mulled over, .web was in limbo.
The IRP was unsuccessful. The IRP panel ruled in October that ICANN had not violated its bylaws in accepting the SCO panel’s decision.
But it gave ICANN a nudge, suggesting that perhaps it could give Vistaprint leave to appeal the original SCO determinations via another mechanism.
In early March, the ICANN board proper decided that:

the Vistaprint SCO Expert Determination is not sufficiently “inconsistent” or “unreasonable” such that the underlying objection proceedings resulting in the Expert Determination warrants re-evaluation.

The board said that the .web/.webs contention set should be processed as normal; in other words: go to auction.
That removed the first barrier to the .web/.webs auction going ahead.
The second barrier was the fact that Visaprint had file two applications for .webs — one regular, one “community”.
By self-identifying as a “community”, Vistaprint qualified for the Community Priority Evaluation. A winning CPE means all competing applications — including the .web applications in this case — would be eliminated.
While the CPE process is far from perfect, I think the chances of Vistaprint winning would be pretty slim.
Perhaps Vistaprint agreed with me. Whatever the thought process, the company has withdrawn its “community” application. The withdrawal was reflected on the ICANN web site at the weekend, according to the little birds at DI PRO.
What this means is that the seven .web applications and Vistaprint’s remaining, non-community .webs application will be going to auction together.
It could be a private auction, where the proceeds are divvied up between the losers, or an ICANN “last resort” auction, where ICANN gets all the money.
Either way, the winning bidder is likely to pay a LOT of cash for their chosen string.
GMO Registry paid $41 million for .shop back in January. I’d be flabbergasted if .web wasn’t eight figures too.
If Vistaprint offers to pay more money for .webs than Web.com wants to pay for .web, Web.com will be eliminated from the race and Vistaprint will get .webs.
In that scenario, the remaining six .web applicants fight it out for control of the gTLD.
However, if Vistaprint loses against Web.com then all of the seven .web applicants fight it out at auction.
Depending on the identity of the winner and the timing of auctions and pre-delegation testing, it could slip into the root and possibly even become available before the end of the year.
That’s assuming no more surprises, of course.
UPDATE: This post originally incorrectly described the rules of the .web/.webs auction. It was updated with a correct explanation at 2120 UTC.

Greimann wins Nominet board seat

Key-Systems general counsel Volker Greimann has been elected to Nominet’s board as a non-executive director, beating two rival candidates.
Nominated by Blacknight and EuroDNS, he got 1,169,785 of the 2,144,612 votes cast, beating Dot Advice CEO Phil Buckingham and Namesco domain development manager Kelly Salter.
Nominet uses a somewhat complex single transferable vote system in its elections, in which members votes are weighted according to how many .uk domains they have under management.
Voting power is capped at 3% of the total pool for each member, so no one registrar or small group of registrars can capture the election.
Salter, who had been nominated by top-ten registrars Go Daddy and LCN.com, was defeated in the first round of voting, with Greimann picking up the majority of the votes as a second preference, enabling him to win.
Buckingham was essentially the “domainer candidate”, backed by Netistrar and Namedropper, who’d promised to address the controversial issue of Nominet’s 50% price increase.
The price increases are arguably less important to registrars which can pass the increase on to their customers, than they are to domainers, which have to swallow the added costs themselves.
Buckingham secured about 34% of the votes in the second round.
Only 15% of the members eligible to vote did so, though that’s up from 12% last year.
The full results can be found here.
Nominet will hold its Annual General Meeting in London tomorrow.

M+M turns $22m profit into $10m loss

Minds + Machines today reported a 2015 loss of $10 million and further outlined its “transformative” restructuring and China strategy.
It’s the second full year of operating results M+M has posted since its first new gTLDs went live, and they’re not encouraging.
Revenue for accounting purposes was $6.3 million, but the cost of sales was $6.2 million, leaving gross profit of just $101,000.
Factoring in $12.1 million of operating expenses, a $7.9 million gain from losing new gTLD auctions, and other expenses, the total loss before tax was $10 million.
That’s compared to the $22 million profit M+M reported for 2014, a number entirely reliant on $33.7 million of auction loss payments.
The company also reported its “billings”, a line item that does not use the accounting method of deferring revenue across the life of a domain and is therefore more in line with incoming cash.
Billings for 2015 were $7.9 million, compared to $5 million in 2014. Gross profit under that measure was $1.7 million, but the $12 million of operating costs still made the company very unprofitable.
Ignoring the auction benefits in 2015, which will not last forever, it’s pretty clear that M+M was a company spending much more operating new gTLDs than it was making from them.
COO/CFO Michael Salazar said in a statement:

However, billings of $7.9 million for the year were simply not of a sufficient scale to cover the associated cost of sales ($6.2 million) and operating expenses ($12.2 million), which combined reached $18.4 million for 2015. Similarly, the $0.6 million savings achieved in the period by the decisions mid-year to stream-line the existing operational set-up were not of a magnitude to have any material impact in the year under review. That said, forfeited cost of sales and operational expenses as a result of the 2015 cost-cutting decisions will amount to $2.7 million in 2016

It’s perhaps little wonder that activist shareholders, apparently not prepared to play the long game, threw out half of the board and key senior executives during the period.
Former PR man Toby Hall took over as CEO in February, replacing co-founder Anthony Van Couvering, and announced earlier this month that M+M is dumping its registrar and back-end registry businesses.
Its registrar customers have been sold to Uniregistry, and it will outsource its registry back-end to Nominet, to save costs.
Salazar said that the two deals will lead to $2 million in savings, but won’t be complete before the fourth quarter. It seems unlikely they’ll have a great impact on 2016 numbers.
Headcount has been reduced from a peak of 61 to 43 at the end of the year, and is expected to drop further to 25. Salazar said this will save it $4.7 million a year.
Even with these cost reductions, M+M will still need to essentially double its revenue in order to hit operating profitability, it seems.
The company is pinning some of its growth hopes on .vip, which it expects to do well in China. It launches May 18.
Hall said in a statement that M+M would not follow the lead of competitors (Famous Four Media springs to mind) by offering first-year registrations for free to build market share. He said:

Based on the enquiries received during Sunrise and feedback gained through our two recent marketing trips to China, it is clear that there is genuine interest in the domain both within and outside of China. As a result, we will not be using a year-one freemium approach to simply inflate year-one registrations. Instead, we intend to be keenly priced to ensure margin to ourselves — and registrations — as well as protect the integrity of the domain. The volume we anticipate to be generated through keen pricing will then support the sales of our premium names in this domain.

The company also plans to invest in its .law sales team, because billings for that gTLD have been behind expectations.
M+M had $34.6 million in the bank and eight outstanding contested new gTLD applications at the end of the year.

Registrars say Amazon is “closing” open gTLD

A group comprising some of the largest domain registrars has claimed Amazon is attempting to close off a new gTLD that it previously indicated would be unrestricted.
The 12-strong group, which includes Go Daddy, Network Solutions and Tucows, also claims that the company’s proposal for a “Registration Authentication Platform” is anti-competitive.
The complaints follow Amazon’s filing of a Registry Services Evaluation Process request with ICANN in March.
The RSEP speaks in broad terms about rejigging the conventional domain registration path so that all .moi sales are funneled through Amazon’s registry site, where registrants will have their eligibility verified and then be offered a set of add-on “technology tools” before being bounced back to their chosen registrar.
Amazon hasn’t said who will be eligible to register .moi domains, nor has it explained what technology tools it plans to offer. I expect the tools will include things such as hosting and security, where many registrars currently make money.
Unsurprisingly, many registrars are not happy about these vague proposals.
In a comment (pdf) to the RSEP filed yesterday, they said:

Ultimately, the use of pre-registration verification and “optional” value added services will negatively impact competition. By tying both practices in a TLD, a TLD Operator can create a “captive audience” via the pre-registration verification and then offering optional services. This will effectively bypass the existing registration and purchase process, putting TLD Operator in a privileged position. The TLD is set up to capture customers earned via the Registrars marketing efforts to promote its own tools and services.

It’s not unusual for “sponsored” or “restricted” gTLDs to implement registry-side verification, they admitted, but said that .moi is meant to be “open”.
They wrote:

While this practice is not explicitly prohibited under gTLDs, we believe that post-delegation inclusion of these practices should only be allowed in compelling circumstances because they are, in effect, retroactively “closing” what was applied for and approved to be operated as an open, generic TLD.

Amazon’s application for .moi, like all of its new gTLD applications, is not entirely clear on what the company’s plans are. There’s vague talk about eligibility, but no details and nothing substantial to suggest a tightly restricted zone.
The signatories to the registrar comment represent the majority of registered domain names. They are: Astutium, Blacknight Internet Solutions, Domain.com, EuroDNS, GoDaddy.com, OpenproviderNetEarth One, Key-Systems, Netistrar, Network Solutions, Nordreg, Realtime Register, Tucows Domains.
One registrar, Com Laude, whose sister company Valideus handles Amazon’s gTLD applications, wrote a comment (pdf) expressing the opposite view.
Com Laude says that it’s not unusual for registries to require registry-side verification. It points to .bank, .pharmacy and .travel as examples.
The company also claims that the 12 registrars are in essence complaining about the idea of vertical integration — where registries and registrars are under common ownership — which is already in place at companies such as Uniregistry and Rightside.
Com Laude’s Jeff Neuman wrote:

We do not believe that it is unacceptable for a company like Amazon to do what these other companies have been doing for some time. To apply different standards to Amazon Registry than it does for each of the other vertically integrated entities would single them out for disparate treatment – especially when there is no factual basis to believe that Amazon Registry has not adhered to its vertical integration-related obligations under the Registry Agreement.

What’s going on here, I suspect, is a bit of a proxy war.
Neither Amazon nor the registrars care a great deal about .moi, I think. The gTLD is merely a canary for Amazon’s 30-odd yet-to-be-launched gTLDs. The company has the rights to potentially more attractive strings, including .book, .song and .tunes.
Amazon originally wanted to make these strings “closed generics”, or what ICANN calls “exclusive access” gTLDs, where only Amazon could register names.
It has since disavowed such plans, but still hasn’t said who will be able to register names in its portfolio or how they will prove eligibility.
.moi was not originally identified as a closed generic by ICANN, but it could represent a model for what Amazon plans to do with the rest of its stable.

.hotel fight gets nasty with “criminal” hacking claims

A group of would-be .hotel gTLD registries have called on ICANN to reject the winning applicant’s bid or be complicit in “criminal acts”.
The group, which includes Travel Reservations, Famous Four Media, Radix, Minds + Machines, Donuts and Fegistry is threatening to file a second Independent Review Process complaint unless ICANN complies with its demands.
Six applicants, represented by Flip Petillion of Crowell & Moring, claim that Hotel Top Level Domain Sarl should forfeit its application because one of its representatives gained unauthorized access to their trade secrets.
That’s a reference to a story we covered extensively last year, where an ICANN audit found that DotBerlin CEO Dirk Krischenowski, or at least somebody using his credentials, had accessed hundreds of supposedly confidential gTLD application documents on ICANN’s web site.
Krischenowski, who has denied any wrongdoing, is also involved with HTLD, though in what capacity appears to be a matter of dispute between ICANN and the rival .hotel applicants.
In a month-old letter (pdf) to ICANN, only published at the weekend, Petillion doesn’t pull many punches.
The letter alleges:

Allowing HTLD’s application to proceed would go agaist everthing that ICANN stands for. It would amount to an acquiescence in criminal acts that were committed with the obvious intent to obtain an unfair advantage over direct competitors.
…
ICANN caught a representative of HTLD stealing trade secrets of competing applicants via the use of computers and the internet. The situation is even more critical as the crime was committed with the obvious intent of obtaining sensitive business information concerning a competing applicant.

It points out that ICANN’s Applicant Guidebook disqualifies people from applying for a new gTLD if they’ve been convicted of a computer crime.
To the best of my knowledge Krischenowski has not been convicted of, or even charged with, any computer crime.
What ICANN says he did was use its new gTLD applicants’ customer service portal to search for documents which, due to a dumb misconfiguration by ICANN, were visible to users other than their owners.
Krischenowski told DI in an emailed statement today:

According to ICANN, the failure in ICANN’s CSC and GDD portals was the result of a misconfiguration by ICANN of the software used (as mentioned at https://www.icann.org/news/announcement-2-2015-11-19-en). As a user, I relied on the proper functioning of ICANN’s technical infrastructure while working with ICANN’s CSC portal.

HTLD’s application for .hotel is currently “On Hold”, though it is technically the winner of the seven-application contention set.
It prevailed after winning a controversial Community Priority Evaluation in 2014, which was then challenged in an Independent Review Process case by the applicants Petillion represents.
They lost the IRP, but the IRP panelists said that ICANN’s failure to be transparent about its investigation into Krischenowski could amount to a breach of its bylaws.
In its February ruling, the IRP panel wrote:

It is not clear if ICANN has properly investigated the allegation of association between HTLD and D. Krischenowski and, if it has, what conclusions it has reached. Openness and transparency, in the light of such serious allegations, require that it should, and that it should make public the fact of the investigation and the result thereof.

The ruling seems to envisage the possibility of a follow-up IRP.
ICANN had told the panel that its investigation was not complete, so its failure to act to date could not be considered inaction.
The ICANN board resolved in March, two days after Petillion’s letter was sent, to “complete the investigation” and “provide a report to the Board for consideration”.
While the complaining applicants want information about this investigation, their clear preference appears to be that the HTLD application be thrown out.

Aussies get to drop the .com

Australia’s ccTLD manager has confirmed that local registrants will be able to register .au domains at the second level, eschewing the usual .com.
auDA said yesterday that its board has approved a plan that will let people register names such as example.au, rather than example.com.au or example.org.au.
The country follows the example of New Zealand and the UK, which have also started to permit second-level registrations in recent years.
auDA agreed with its policy committee that direct .au registrations would:

– make available domain names which are shorter, more appealing and more memorable
– give Australians more choice in deciding what domain name to register
– respond to market demand
– be more attractive to natural individuals than the current option, id.au
– strengthen the “.au brand” in a globally competitive market
– add value to all three main categories of users – registrars and resellers, registrants and ultimate users of the .au domain name system.

There’s no timeline yet on when 2LDs will become available, much less a policy on how potential conflicts will be handled, but auDA said it will provide updates later in the year.

.sucks “gag order” dropped, approved

Vox Populi, the .sucks registry, has had controversial changes to its registrar contract approved after it softened language some had compared to a “gag order”.
ICANN approved changes to the .suck Registry-Registrar Agreement last week, after receiving no further complaints from registrar stakeholders.
Registrars had been upset by a proposed change that they said would prevent brand-protection registrars from publicly criticizing .sucks:

The purpose of this Agreement is to permit and promote the registration of domain names in the Vox Populi TLDs and to allow Registrar to offer the registration of the Vox Populi TLDs in partnership with Vox Populi. Neither party shall take action to frustrate or impair the purpose of this Agreement.

But Vox has now “clarified” the language to remove the requirement that registrars “promote” .sucks names. The new RRA will say “offer” instead.
Registrars had also complained that the new RRA would have allowed Vox to unilaterally impose new contractual terms with only 15 days notice.
Vox has amended that proposal too, to clarify that changes would come into effect 15 days after ICANN has given its approval.
Vox CEO John Berard told ICANN in a March 18 letter:

VoxPop’s intent was never to alter any material aspect of the Registry Registrar Agreement. Our intent was to clarify legal obligations that already exist in the Agreement, and conform the timeframes for any future amendments with those specified in our ICANN registry contract.

Did the DotConnectAfrica judge make a big dumb mistake?

The court ruling that granted DotConnectAfrica a preliminary injunction preventing ICANN delegating .africa seems to be based to a large extent on a huge error by the judge.
In explaining why he was allowing DCA v ICANN to proceed, despite DCA’s signing away its right to sue when it filed its new gTLD application, California district judge Gary Klausner seems to have confused DCA with rival .africa applicant ZACR.
In his Tuesday ruling, Klausner said that evidence supports the claim that ICANN was determined to flunk DCA’s application no matter what.
The key evidence, according to the judge, is that the Initial Evaluation of DCA’s application found that it did have enough support from African governments to pass its Geographic Names Review, but that ICANN subsequently reversed that view in Extended Evaluation.
He wrote:

DCA claims that “the process ICANN put Plaintiff through was a sham with a predetermined ending – ICANN’s denial of Plaintiff’s application so that ICANN could steer the gTLD to ZACR.”
…
In support, DCA offers the following evidence. ICANN’s initial evaluation report in July 2013 stated that DCA’s endorsement letters “met all relevant criteria in Section 2.2.1.4.3 of the Applicant Guidebook.” (Bekele Decl. ¶ 40, Ex. 27, ECF No. 17.) After the IRP Decision, ICANN performed a second evaluation on the same information originally submitted by DCA. In the second evaluation, however, ICANN found that the endorsement letters did not meet the same criteria applied in the first evaluation

He later writes:

Despite ICANN’s contention, the evidence presents serious questions pointing in favor of DCA’s argument. First, a March 2013 email from ICC to ICANN stated that ICANN needs to clarify AUC’s endorsements since AUC properly endorsed both DCA and ZACR. (Bekele Decl. ¶ 30, Ex. 19, ECF No. 17.) Subsequently, ICANN’s July 2013 initial evaluation report found that the endorsement letters have “met all relevant criteria in Section 2.2.1.4.3 of the Applicant Guidebook.” (Bekele Decl. ¶ 40, Ex. 27, ECF No. 17.) Because ICANN found DCA’s application passed the geographic names evaluation in the July 2013 initial evaluation report, the Court finds serious questions in DCA’s favor as to whether DCA’s application should have proceeded to the delegation stage following the IRP Decision.

The document “Bekele Decl. ¶ 40, Ex. 27” referred to is exhibit 27 of DCA CEO Sophia Bekele’s March 1 declaration, filed in support of its preliminary injunction motion.
The problem is that that exhibit is not the Initial Evaluation report for DCA’s .africa bid, it’s the IE report for rival ZACR (aka UniForum).
Read it here (pdf).
DCA’s own application never received a scored IE report. At least, one was never published.
It only got this (pdf), which states simply “Overall Initial Evaluation Summary: Incomplete”. That document is dated July 3, 2013, almost two weeks before the ZACR report.
Bekele’s declaration even states that exhibit 27 is the IE report for the ZACR application.
It’s not clear to this non-lawyer how important this pretty basic error is to Klausner’s thinking, but as a layman it looks pretty crucial.
It certainly seems like something that needs to be addressed, given that the apparent misunderstanding plays into both the decision to allow the lawsuit to proceed and the decision that DCA’s complaint may have merit.
Several other exhibits cited in the ruling — including emails from the InterConnect Communications evaluators who carried out the Geographic Names Review — have been redacted by the court.
It’s possible there are smoking guns contained within these censored documents that were more influential on the ruling.
It’s also notable that ICANN is continuing to redact the court documents it publishes on its web site, beyond those filed under seal and censored by the court.

Afilias goes it alone with .green as DotGreen bows out

DotGreen Community has withdrawn from its partnership with Afilias, leaving the registry operating the .green gTLD solo.
The new management means that renewal prices for sunrise names will be slashed, and the size of the premium names list will be reduced.
DotGreen was originally a .green gTLD applicant, but it withdrew its application before the auction, which Afilias subsequently won.
It made a deal with Afilias to run sales and marketing for the domains, with Afilias handling all the technical stuff. The plan was to use profits to help environmental causes.
But Afilias told its registrars in an email this week:

Unfortunately, DotGreen Community has informed us that it is no longer able to discharge these responsibilities, and has turned the sales and marketing responsibilities back over to Afilias.

The gTLD has a new logo and a new website at get.green.
Afilias said it will no longer charge a extra $50 for renewals of sunrise period names.
It also plans to revise its premium names list by November, and will probably reduce the size of the list, releasing names at regular registry prices.
.green domains haven’t exactly been flying off the shelves with DotGreen running marketing.
It went into general availability in March last year but only has about 2,200 names in its zone file.

No, .kids isn’t a community either

DotKids Foundation has comprehensively lost is .kids Community Priority Evaluation.
The company’s CPE results came out at the weekend (pdf), showing a score of 6 out of the 16 available points, a long way short of the 14-point passing score.
Like other “community” new gTLD bids before it, .kids failed because the Economist Intelligence Unit panel decided that the application was an attempt to create a community rather than represent an existing one. It wrote:

The Panel determined that this application refers to a “community” construed to obtain a sought-after generic word as a gTLD string, and that the application is attempting to organize the various groups mentioned in the documentation through a gTLD.

The application scored a big fat 0/4 on the question of whether the community exists and, as a knock-on effect, another 0/4 on whether the .kids string represents the community.
It picked up 3/4 for its registration policies and 3/4 on community endorsement.
The CPE failure means DotKids will have to face rival Amazon at auction, where one imagines the not-for-profit foundation will have a hard time winning.
ICANN’s CPE pipeline currently only has one active application, where Merck KGaA is fighting to avoid an auction with rival Merck Registry Holdings, Inc.
The latter .merck, and Vistaprint’s .webs application, have both also been invited to CPE.